PathologyLabTraining

Root Cause Analysis and CAPA in Pathology: Investigating Non-Conformities the Right Way

12 min read · Published: 2025-01-08 · Updated: 2026-06-08

When something goes wrong in the laboratory — a sample lost, a result transcribed incorrectly, an analyser drifting out of specification — the instinct is to fix the immediate problem and move on. Under ISO 15189:2022 and modern NHS patient-safety expectations, that is not enough. A competent biomedical scientist (BMS) must be able to investigate why the non-conformity happened, design corrective and preventive actions (CAPA) that stop it recurring, and then prove those actions actually worked. This article walks through the methodology UK assessors and patient-safety teams expect to see.

Why Investigation Matters: The ISO 15189 and NHS Framework

ISO 15189:2022, the international standard against which UKAS (the United Kingdom Accreditation Service) assesses medical laboratories, deals with non-conforming work and corrective action across two main areas. Clause 7.5 (Nonconforming work) requires the laboratory to identify, evaluate the clinical significance of, and control any work that does not meet its own procedures, quality specifications or user requirements. Clause 8.7 (Nonconformities and corrective action) then requires the laboratory to respond to the non-conformity, determine its cause(s), implement corrective action proportionate to the effect, and review whether that action was effective. Clause 8.6 (Improvement) closes the loop by feeding learning back into the quality management system.

The standard takes a risk-based approach throughout: your response should be proportionate to the potential for patient harm. Alongside accreditation, NHS laboratories operate within national patient-safety arrangements. The Learn from Patient Safety Events (LFPSE) service has replaced the old National Reporting and Learning System (NRLS) for recording events, and the Patient Safety Incident Response Framework (PSIRF) has replaced the former Serious Incident Framework as the NHS approach to responding to and learning from incidents. Where transfusion is involved, separate statutory and professional schemes apply (see below).

The common thread across all of these is a deliberate move away from blaming individuals and towards understanding the system that allowed the error to occur.

Correction vs Corrective Action: The Distinction Examiners Test

This is the single most commonly confused point in quality management, and UKAS assessors raise it repeatedly. The terms are not interchangeable.

| Term | Definition | Pathology example | |------|-----------|-------------------| | Correction | An immediate action to deal with the detected non-conformity itself — containment | Recalling and re-testing the affected sample; issuing a corrected report; quarantining a reagent lot | | Corrective action | Action to eliminate the cause of the non-conformity so it cannot recur | Redesigning the worklist, retraining staff, adding an LIMS validation rule, changing the SOP | | Preventive action | Action to eliminate the cause of a potential non-conformity before it ever happens | Risk-assessing a new analyser interface and building in safeguards before go-live |

A correction treats the symptom for one patient. A corrective action treats the disease so no future patient is affected. If your investigation produces only a correction — "the wrong result was amended and the clinician informed" — you have done half the job and an assessor will raise a finding. Note also that ISO 15189:2022 folds the old idea of "preventive action" into risk-based thinking that runs through the whole management system, but the concept of preventing recurrence remains central.

Step One: Containment and Immediate Response

Before any deep investigation, control the situation. Working through ISO 15189:2022 Clause 7.5, the immediate actions are:

1. Halt and contain. Stop the affected examination process and withhold reports where there is any risk of harm to patients. 2. Assess clinical significance. Evaluate the impact on results that have already been released, or could have been released, before the problem was found. 3. Escalate. Inform the appropriate person — typically the laboratory manager, quality lead and, where patients are affected, the requesting clinician. 4. Document. Raise a non-conformity record in your quality system at the point of discovery, not retrospectively. 5. Decide on acceptability. Determine whether any nonconforming work can be released with caveats, must be repeated, or must be withdrawn.

Only once the patient-safety position is secured should you turn to root cause analysis. Containment buys you the time to investigate properly rather than reacting under pressure.

Step Two: Root Cause Analysis with the 5 Whys

Root cause analysis (RCA) is a structured method for moving past the obvious surface cause to the underlying system failure. The 5 Whys is the simplest tool and works well for single, relatively linear problems.

Starting from the problem statement, ask "why?" repeatedly — typically around five times, though the number is not fixed — until you reach a cause that is within the laboratory's control to change. A worked example:

The root cause here is not "the BMS missed the flag" (a person) but "the system presents critical flags in a way that is easy to miss" (a design). That distinction is what separates an effective corrective action from a useless one such as "told staff to be more careful."

Common pitfalls with the 5 Whys:

Step Three: Fishbone (Ishikawa) Analysis for Complex Problems

When a non-conformity has several contributing causes — which most serious incidents do — a fishbone (Ishikawa) diagram is more powerful than a single chain of whys. The problem sits at the "head" of the fish, and contributing causes are grouped along the "bones." A common categorisation for laboratory work, adapted from the classic 6Ms, is:

You brainstorm causes under each heading, then test which ones genuinely contributed using your records — IQC charts, maintenance logs, training records, LIMS audit trails. The strength of the fishbone is that it forces a systems view and surfaces latent conditions, which aligns with the NHS PSIRF emphasis on understanding how parts of the system interact rather than isolating one individual's mistake. For PSIRF learning responses specifically, NHS England promotes the Systems Engineering Initiative for Patient Safety (SEIPS) framework, which examines the work system (people, tasks, tools and technology, internal and external environment, and organisation) in the same systems-thinking spirit.

Step Four: Designing Effective CAPA

Once the genuine root cause(s) are identified, design actions that eliminate them. Effective CAPA shares several characteristics:

Distinguish corrective actions (stopping recurrence of this problem) from genuinely preventive ones (scanning for the same latent failure elsewhere). If a flag-visibility problem caused a potassium error in biochemistry, a preventive action asks whether the same display weakness exists for critical flags in haematology or coagulation.

Step Five: Verifying CAPA Effectiveness and Closing the Loop

This is the step most often missed, and it is explicit in ISO 15189:2022 Clause 8.7: the laboratory must review the effectiveness of any corrective action taken. Implementing an action is not the same as proving it worked.

Verification approaches include:

Only when the evidence shows the action has eliminated the cause should the non-conformity record be formally closed. Record the verification evidence, not just a "closed" status — UKAS assessors look for a functioning closed-loop system: detect, investigate root cause, act, and verify. An unverified corrective action is itself a common finding. Finally, feed the learning into the wider quality management system under Clause 8.6 (Improvement), and where appropriate share it with other disciplines so a single investigation prevents multiple future incidents.

External Reporting: When the Loop Extends Beyond the Lab

Some non-conformities must be reported externally as well as managed internally:

External reporting never replaces your internal RCA and CAPA; the two run in parallel, and a national report is only as good as the local investigation that informs it.

Frequently Asked Questions

What is the difference between a correction and a corrective action?

A correction is the immediate fix for the problem in front of you — re-testing a sample, amending a report, quarantining a reagent. A corrective action eliminates the underlying cause so the problem cannot recur, for example changing an SOP, adding a LIMS validation rule or redesigning a worklist. ISO 15189:2022 requires both, and an investigation that produces only a correction is incomplete.

How many "whys" should a 5 Whys analysis have?

Five is a guideline, not a rule. You keep asking "why?" until you reach a cause that the laboratory can realistically change — sometimes that takes three steps, sometimes seven. The test of a good root cause is that it points to a system or process change rather than to an individual being told to "be more careful."

When should I use a fishbone diagram instead of the 5 Whys?

Use the 5 Whys for simple, single-cause problems with a fairly linear chain of events. Use a fishbone (Ishikawa) diagram when an incident has several contributing causes across people, process, equipment, materials, measurement and environment — which is true of most serious incidents. The fishbone forces a systems view and is well suited to the NHS PSIRF and SEIPS emphasis on understanding the whole work system.

Does ISO 15189:2022 still require "preventive action"?

The 2022 edition embeds preventive thinking within risk-based management that runs across the whole standard, rather than as a separate clause. The concept absolutely remains: you are expected to identify and act on potential non-conformities before they occur, and to check whether a root cause found in one area exists elsewhere in the laboratory.

How do I prove a corrective action actually worked?

Verification is required by Clause 8.7. Re-audit the process, monitor the relevant quality indicator over a defined period, review IQC and EQA performance, and observe the revised process in practice. Only close the non-conformity once objective evidence shows the cause has been eliminated, and keep that evidence on record for assessment.

Why does the NHS discourage blaming individuals in incident investigations?

Blame stops people reporting errors and rarely fixes the underlying problem, because most errors arise from system and process weaknesses rather than individual carelessness. PSIRF and a systems-thinking approach focus on understanding how the work system contributed, which produces stronger, more durable corrective actions and a more open reporting culture.

Further training

Root cause analysis and CAPA sit at the heart of laboratory quality and are tested in IBMS portfolios, OSPEs and band promotion interviews. Build your wider quality knowledge across the NHS Laboratory Training hub and these closely related guides:

Related Articles